Crosscoder
- model_A: google/gemma-2-2b-it
- model_B: google/gemma-2-2b-jpn-it
- d_sae: 2**15
- hook-point: blocks.14.hook_resid_pre
- Total training tokens (1B):
  - monology/pile-uncopyrighted: 0.5B tokens (tokenizer: model_A)
  - statmt/cc100 (ja): 0.5B tokens (tokenizer: model_B)
- Training details:
Implemented from https://github.com/ckkissane/crosscoder-model-diff-replication/tree/main

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for haggingfacehyz/crosscoder-gemma-2-2b-it-gemma-2-2b-jpn-it-L13

Base model

Finetuned

Finetuned

(724)

this model